特邀报告

特邀报告

宋柔
教授
广东外语外贸大学
张民
教授
苏州大学

标题:Semantics, Discourse and Machine TranslationSlides1下载

摘要:

Semantics in linguistics refers to the meanings of linguistic expressions and their relations. Discourse in linguistics refers to a unit of language longer than a single sentence. They are attracting more and more attentions in the research community of computational linguistics and NLP in very recent years. This talk consists of two parts, i.e., 1) semantics and machine translation, 2) discourse and machine translation. We will first give an overview about semantics/discourse and review the research state-of-the-art of semantics/discourse from both linguistics and computational viewpoints, and then discuss how machine translation can benefit from semantics/discourse-level information. In particularly, we will report the research effort of Chinese Micro-topic structure theory pioneered by Prof. Song Rou. Finally, we conclude the talk with some future direction discussions.

语言学上的语义指的是语言单元的意义和它们之间的关系。语言学上的篇章指的是大于一个单句(或汉语小句)的语言单位。篇章是由多个句子(或汉语小句)组成、结构和语义完整、功能明确的语言单位。作为交际单位,它是一个完整的语义统一体,不是多个句子的无序堆砌和简单组合,而是要表达一个完整的语义、思想和意图。语义和篇章在计算语言学和自然语言处理领域正得到越来越多的关注。这个报告包括两部分:语义和机器翻译;篇章和机器翻译。本报告将重点介绍宋柔老师提出的面向汉语微观话题结构的汉语广义话题结构的理论和实践,以及基于该理论和实践的机器翻译方法。最后,本报告给出相关研究的未来发展方向。

简历:

SONG Rou(宋柔): a professor and Ph.D. supervisor at Applied Linguistics and Computer Application in Beijing Language and Culture University, received his Bachelor degree in mathematics and mechanics from Beijing University in 1968 and his mater Master degree in computer science from Beijing University in 1981. He has been working on Chinese Information Processing study for tens of years as the PIs of more than 10 national-level projects with the research focuses on discourse analysis, Chinese word segmentation, Computer-aided proofreading, Chinese word attribute, Chinese Orthographic Computing and Chinese POS and so on. He has published more than 100 papers at leading journals and conferences in computer science and linguistics. He has developed and commercialized several softwares with two patents. He has received several awards from Beijing City and MOE, China. He has been appointed as guest professors in a few domestic and oversea universities and research institutes.

ZHANG Min(张民): a distinguished professor and vice dean of the school of computer science and technology, director of the research Institute for Human Language Technology at Soochow University (China), received his Ph.D. degree in computer science from Harbin Institute of Technology (China) in 1997. He has studied and worked oversea in industry and academy since 1997 to 2013. His current research interests include machine translation and natural language processing. He has co-authored 2 Springer books and more than 130 papers in leading journals and conferences, and co-edited 13 books published by Springer and IEEE. He is an associate editor of IEEE T-ASLP (2015-2017).

吕正东
博士
华为诺亚方舟实验室

标题:Recent Progress on Neural Machine Translation at Noah's Ark LabSlides下载

摘要:

I will introduce the recent progress made on neural machine translation (NMT) at Noah's Ark Lab. More specifically, I will talk about 1) the memory-based deep architecture for machine translation, and 2) neural machine translation as a classification problem. For 1), I will discuss the new deep architecture based on recursive read-write with external unbounded memory, with state-of-the-art performance on Chinese=English translation. For 2), I will introduce a new neural machine translation framework consisting purely of convolutional neural network components.

我会介绍华为诺亚方舟实验室在神经机器翻译(NMT)方向上的最新进展。这些进展分为两个方面,1)基于外置记忆的深度结构,和2)神经机器翻译作为一个分类问题。关于1),我将会主要介绍我们的一种基于外置记忆的反复读写的深度架构,以及这个架构在机器翻译上的应用。这个深度架构在中文-英文的翻译任务上有很好效果。关于2), 我将会介绍一个新的完全基于卷积神经网络的神经机器翻译框架。

简历:

Zhengdong Lv is a senior researcher at Noah's Ark Lab, Huawei. Zhengdong Lu got his Ph.D degree from Oregon Health & Science University. Before joining Huawei, Dr. Lu worked as a postdoctoral researcher in University of Texas at Austin and an associate researcher in MSRA.

The Noah's Ark Lab is a research lab of Huawei Technologies, located in Hong Kong and Shenzhen. The mission of the lab is to make significant contributions to both the company and society by innovating in data mining, artificial intelligence, and related fields. Founded in 2012, the lab has now grown to be a research organization with many significant achievements in both academia and industry.

青年学者特邀报告

黄书剑
博士,助理研究员
南京大学

标题:关于机器翻译中的参数学习问题的思考Slides下载

摘要:

传统的统计机器翻译系统都采用对数线性模型融合语言模型、翻译模型等子系统,并采用最小错误率训练的方法对对数线性模型的参数进行学习。近年来,很多研究者注意到这一参数学习框架存在一些问题,如难以融入大量的特征,模型表达能力有限等。本报告介绍一些机器翻译参数学习的研究进展,力图分析一些仍然未解决的问题,并探讨可能的解决方案。

简历:

黄书剑,博士,主要研究方向为自然语言处理、机器翻译。2002年进入南京大学计算机科学与技术系学习,于2006年和2012年分别获得学士和博士学位,在读期间曾于微软亚洲研究院、美国卡耐基梅隆大学进行长期学术访问。博士毕业后留校任助理研究员,主持和参与多项国家及江苏省自然科学基金研究项目,并与相关企业单位开展科研合作。发表文章十余篇,包括ACL、EMNLP、CoNLL等。联系方式:huangsj@nju.edu.cn。

毛存礼
博士,硕士生导师
昆明理工大学

标题:东南亚语言信息处理及机器翻译研究Slides下载

摘要:

在国家“一带一路”战略下,为推动与东南亚国家的交流与合作,研究东南亚语言的机器理解以及实现汉语—东南亚语言的机器翻译有着非常重要的意义。然而,由于东南亚语言属于低资源语言,汉语—东南亚语言双语平行语料非常稀缺,并且在词法及句法方面东南亚语言与汉语存在较大差别。本报告将介绍越南语、老挝语、柬埔寨语、泰国语、缅甸语等东南亚国家小语种语言信息处理的研究现状,汉语-东南亚语言双语知识库、语料库构建,东南亚语言信息处理、跨语言检索及机器翻译翻译等研究工作。

简历:

毛存礼,博士,硕士生导师。中国中文信息学会青年工作委员会委员,2014年在昆明理工大学获得博士学位。 2005年至今就职于昆明理工大学信息工程与自动化学院,昆明理工大学智能信息处理重点实验室。主要研究方向为自然语言处理、信息检索及机器翻译等。近年来在Pattern Recognition Letters、软件学报等期刊和国际会议发表论文10余篇。分别获得云南省科技进步一等奖、云南省自然科学二等奖、云南省科技进步三等奖各1项。

苏劲松
博士,副教授
厦门大学

标题:一种面向统计机器翻译的上下文感知主题模型Slides下载

摘要:

译文词汇选择是统计机器翻译的重要任务。传统研究主要致力于使用句子层次信息的局部上下文或者篇章层次的主题上下文来改善译文词汇选择,然而却忽视了两种上下文信息之间的相互联系。在这个工作中,我们提出一种面向统计机器翻译译文词汇选择的上下文感知主题模型。该模型不仅建模了局部上下文和全局上下文对译文词汇选择的影响,并且考虑了两者之间的相互影响。具体而言,我们在建模过程中将目标译文词汇作为隐变量,以此来建立局部上下文和全局上下文之间的联系。此外,我们引入吉布斯采样方法来进行模型的参数训练和推导。基于该模型,我们对翻译文本进行推导并获得一个新的翻译概率,该概率可以融入对数线性模型以改善译文词汇选择。NIST汉英翻译的实验结果和分析证明了模型的有效性。

简历:

苏劲松,现厦门大学副教授。2007年毕业于厦门大学,获得工学硕士学位,2011年毕业于中国科学院计算技术研究所,获得工学博士学位。主要研究方向是自然语言处理,机器翻译。近年来共发表学术论文10余篇,多篇发表于ACL,EMNLP,COLING等自然语言处理重要会议。

中国中文信息学会 中科院合肥研究院 中科院合肥智能所 科大讯飞 中译语通 沈阳雅译
版权所有:中国科学院合肥物质科学研究院,智能机械研究所
Copyright CWMT 2015. All rights reserved